摘要: |
An information fusion fault diagnosis approach is proposed for vehicle fault diagnosis in this paper.It is based on RBF neural network for data level fusion diagnosis and D-S evidence theory for decision level fusion diagnosis.Using for coolant temperature sensor,oxygen sensor,manifold absolutely pressure sensor ageing fault,the results show that with the increase of input source number,fusion diagnosis precision is improved; And with RBF neural network diagnosis as a source of evidence for decision fusion,the precision of the fusion diagnosis is improved up to 90%,and the reliability of the final diagnosis results is also increased. |